As mentioned in [[@sedleyIntegrativeTinnitusModel2016|my notes from a paper]], it seems tinnitus is a result of gradual shift of prior of how loud is "silence". This can shift can get triggered if for some reason acoustic noise from sub-auditory reasons increases in precision (or decreases in variance) or increases in mean, making resulting posterior (that combines prior of silence) to be louder than complete silence. Gradually these prediction errors force top-down expectations into auditory cortex (prior) to be louder than before (these top-down expectations are probably encoded in memory in para-hippocampus). Because this increased loudness of prior in the brain seems like a subtle signal, it's unlikely to be fixed/corrected using broad based techniques such as transcranial electric stimulation [[Overview of brain input-output techniques#tES]]. This subtle effect is also hard to detect using modern brain imaging methods, so no wonder we're yet to find a neural correlate of tinnitus across hearing-level matched controls. It is generally believed that tinnitus has two aspects: - Loudness - Annoyance And these two _could_ potentially be separate factors experienced by a person and also are potentially modulated by different networks in the brain. ![[Screenshot 2021-11-24 at 11.09.24 AM.png]] (via [here](https://www.youtube.com/watch?v=Pf9s8auwBj4))